ScaleDB extends the MySQL database to manage Streaming and Real-Time Data. The database extensions use special stream engine technologies that integrate with the existing database and storage kernel. This approach adds specialized stream features (including management of Time Series Data) to the existing database functionality such that stream data can be managed by the database while maintaining complete and unified database functionality.
ScaleDB transforms a single database instance (such as MySQL or MariaDB) to a cluster of database and storage services that provides scalability and high availability features that exceed the capabilities of a single database instance. A ScaleDB cluster is made of two tiers – a storage tier and a database tier. The database tier includes multiple MySQL instances that process shared data that is managed in the storage tier. Both tiers use commodity or virtual machines. For Streaming and Time-Series Data ScaleDB provides the following:
Velocity – A ScaleDB cluster ingests millions of inserts per second.
Volume – ScaleDB is a disk based solution. The data persists on the disks of the machines in the storage tier. A ScaleDB cluster is tuned to process Streaming and Time-Series Data such that data volume does not impact performance.
Query functionality – ScaleDB executes Business Intelligence (BI) types of queries over Streaming and Time-Series Data with “pushdown” technology. With the pushdown technology, queries are “pushed” from the database tier to the storage tier and are executed next to the data (similar to MapReduce). This approach allows for distributed processing where billions of rows are evaluated within few seconds.
Query rewind – As Streaming and Time-Series Data is unbounded, ScaleDB offers a rewind mechanism where queries are executed in a sequence of cycles to provide continues view on incoming data. This approach allows applying a SQL query to unbounded data stream.
Simplicity – Streaming and Time-Series Data is processed with the same ease and together with other data. The schema is described using SQL through one of the database nodes in the cluster and becomes available to use to all the nodes in the cluster. Scaling is done by adding nodes to the cluster without the need to redesign the schema or partition/shard the data. High Availability (HA) is transparent.
TCO – ScaleDB offers the best TCO – it is simple to develop, deploy and manage. It provides the highest level of performance and scalability and uses commodity hardware. It harnesses the MySQL echo system to support the most complex Streaming and Time-Series Data applications. ScaleDB combines both stored and streaming/time-series data in an integrated MySQL environment offering large scale real-time streaming applications Velocity and Volume at a very low TCO, enabling a new host of critical business operation analytics implementations.